Time Travel

Sifting through the report from the Berlin Police about bike accidents (see my yesterday post), there are some remarks that I never considered consciously. I usually try to avoid busy hours when I am using a bike. I find it less stressful and I can relax accordingly. When I started to regularly spend time on the bike, I was living in Istanbul, and I did most of my leisure rides early in the morning (before 7.30 am) to be at home when the city gets congested. I probably implicitly felt that it was safer, but I did not know about it.

The “Sonderuntersuchung” about bike accidents in Berlin addresses the time of the accidents (p. 2 from the report 2020). The report highlights that the number of bike traffic accidents depends on the season (more accidents occur during spring and summer when more people are cycling), and the time when accidents happens matches the overall pattern of road accidents (car or motorcycle accident). My guess is that it is (obviously) highly correlated with traffic intensity, but I did not see some number about traffic intensity in the first pages of the report.

Picture of the day

This is the code and the graph of the hour of bike crashes split by the day of the week for Berlin and Brandenburg.

library(sf)

# Select Berlin and Brandenburg
# ULAND = 11 (Berlin) | 12 (Brandenburg)
berbrand <- read.csv2(pipe("awk 'BEGIN {FS=\";\"} {if ($3 == 11 || $3 == 12 ) print $0}' Unfallorte2020_LinRef.csv"),
header=FALSE,
colClasses = c(rep("character", 2),
               rep("factor", 18),
               rep("character", 4),
               rep("factor", 1)
))
readLines("Unfallorte2020_LinRef.csv", n=1) |>
  strsplit(";") |> unlist()  |> tolower() -> colnames(berbrand)
colnames(berbrand)
levels(berbrand$uwochentag) <- c("So", "Mo", "Di", "Mi", "Do",
                                 "Fr", "Sa")

berbrand$uwochentag <-
  factor(berbrand$uwochentag,levels(berbrand$uwochentag)[c(2:7,1)])

# Colour palette
rygb <- c("blue", "yellow", "red", "yellow", "blue")
color.palette <- colorRampPalette(rygb)
color_palette <- color.palette(24)

head(berbrand)
library(ggplot2)
library(dplyr)

jpeg("bike_unfaelle_stunde.jpg", width=800, quality=90)
ggplot(berbrand, aes(x = uwochentag, group=ustunde)) +
  geom_bar(stat = "count",
           aes(fill = ustunde), position = "dodge") +
  stat_count(geom = "text",
             aes(label = ustunde),
             position=position_dodge2(1),
             colour = "black", size = 2.5, vjust=1) +
                      theme_minimal()+ guides(fill="none") +
                      scale_fill_manual(values = color_palette) +
                      labs(x = "Wochentag")
dev.off()

Accident involving a bike by weekday and hour. Data Copyright: DL-DE BY 2.0 Statistische Ämter des Bundes und der Länder, Deutschland, 2021.

Overall, during the week, accidents occur between 7am and 9pm and the peak is situated between 3pm and 6pm. Each weekday has almost the same traffic accident rate, but also a similar pattern. In contrast, the weekend records fewer cases, but proportionally more during the night.